CODES AND STANDARDS
ratio and any presence of transverse reinforcement. The Part 8: Slump flow test for self-compacting concrete
values for such corrections are currently given in the UK
National Annex (NA) to BS EN 13791 but will form part
of the soon-to-be-published revised BS EN 12504-1 with a
UK National Annex. In this Annex there is a requirement
to report both the measured strength and the corrected
strength. In addition, it is useful and sometimes essential to
know the voidage of the cores for the interpretation of the
information and the UK NA requires the excess voidage to
be measured and reported. It should be noted that the cor-
rected core strength does not include a correction for void-
age as such a correction is inappropriate when determining
characteristic in-situ strength.
The revised BS 6089 is intended to provide guidance
on planning an investigation, selection of test methods,
selection of the test location, assessment of individual core
results within a group and assessment where the strength
of concrete based on test specimens is in doubt. It will also
provide guidance on aspects not covered by BS EN 13791
such as: the assessment of an unknown structure using a
margin based on the t-statistic; use of indirect methods
without correlation to core strength, relative testing, ie,
comparison of a volume of concrete under investigation
with concrete in similar elements that has been accepted;
action when the producer has declared non-conformity.
prEN 206-9 Concrete Part 9
BS EN 12350 Parts 812
A new draft part to BS EN 206(7) Concrete has been passed
to CEN. Part 9 there are no Parts 28 Additional rules
for self-compacting concrete will incorporate additional
rules relating to the use of five test methods for self-com-
pacting concrete. This part will eventually be merged into
the main BS EN 206-1 when it is revised in 2010. The five
test methods it refers to for assessing self-compacting con-
crete will be covered in BS EN 12350(8) Testing fresh con-
crete as follows:
BRE Digest 330: Alkali-silica reaction in
concrete the case for revision Part I
Any change to The current BRE Digest 330 was
guidance that published in 2004(1) updating the
previous edition published in 1999(2).
has worked so
The main change was the incorporation
successfully for so
of BRE Information Paper 1/02(3)
many years must introducing guidance in the application
be undertaken of metakaolin, silica fume and lithium
with considerable salts with other minor editorial revisions.
care that short-
term expediency PAUL LIVESEY*
does not result
in longer-term
problems. T he first guidance documents on ways of minimising
the risk of damage due to alkali-silica reaction in con-
crete were published in the early 1980s by the Concrete
Society (Technical Report 30)(4) and BRE Digest 330.
There has been no reported verified case of deleterious
alkali-silica reaction in UK concrete where this guidance
has been implemented. In fact, as the accompanying illus-
trations testify, some exceptional concrete structures have
been built successfully using the guidance.
The case for review
It is five years since the review leading to the current edi-
42 JULY 2009 CONCRETE
CONCRETE July 09 33-48.indd 42
Part 9: V-funnel test for self-compacting concrete
Part 10: L-box test for self-compacting concrete
Part 11: Sieve segregation test for self-compacting
concrete
Part 12: J-ring test for self-compacting concrete
The present position with the above draft Standards is
that the drafts have gone through CEN enquiry stage and
received a positive vote. All the comments received (tech-
nical and editorial) have been considered and the agreed
amendments forwarded to CMC via DIN in April. We
are awaiting publication of the revised draft standards
in English, French and German in order to go to the for-
mal vote stage, probably sometime between June and
September. Provided a positive vote is received, only edi-
torial comments will be considered before final publication
in late 2009/early 2010.
References:
1. BRITISH STANDARDS INSTITUTION, BS EN 13670.
Execution of concrete structures. BSI, London, draft.
2. BRITISH STANDARDS INSTITUTION, BS EN 1992-1-1.
Eurocode 2: Design of concrete structures. BSI, 2004.
3. BRITISH STANDARDS INSTITUTION, BS 6089. Guide to
assessment of concrete strength in existing structures.
BSI, 1981.
4. BRITISH STANDARDS INSTITUTION, BS EN 13791.
Assessment of in-situ compressive strength in structures
and pre-cast concrete components. BSI, 2007.
5. CONCRETE SOCIETY. Concrete core testing for strength.
Technical Report 11, The Concrete Society, Camberley,
1987.
6. BRITISH STANDARDS INSTITUTION, BS EN 12504-1.
Testing concrete in structures. Cored specimens. Part 1 -
Taking, examining and testing in compression. BSI, 2000.
7. BRITISH STANDARDS INSTITUTION, BS EN 206-1.
Concrete. Specification, performance, production and
conformity. BSI, 2000.
8. BRITISH STANDARDS INSTITUTION, BS EN 12350.
Testing fresh concrete. BSI, 2009.
tion and ten years since the last major review. During that
time, there have been changes:
to the concrete industry approach to sustainability and
use of recycled materials, responding to changing
attitudes by clients, UK and EU governments
in concrete technology with more widespread use of
self-compacting and high performance concretes
requiring novel admixtures
in cement production with changes to the product mix
brought by pressure to reduce embodied energy and
carbon dioxide
in all constituent material production to reduce waste
in manufacture resulting in changes to product
characteristics
increasing EU regulatory involvement in safety,
durability and fitness for purpose in construction.
Any change to guidance that has worked so success-
fully for so many years must be undertaken with consid-
erable care that short-term expediency does not result in
longer-term problems.
The original guidance was drafted on the strength of
a considerable volume of laboratory and site investiga-
tion evidence. Since the late 1980s, the volume of such
laboratory investigations has dwindled to almost nothing.
While petrography analysis of cores from existing struc-
23/06/2009 12:43:45

tures is frequently undertaken, commercial confidentiality
has limited the visibility of findings. In this situation it is
difficult to make cases for changes in guidance based on
new evidence. However, a number of the original criteria,
although based on laboratory and site evidence at the time,
were deliberately conservative bearing in mind the limited
experience in their application. Re-examination of these
criteria in the light of over twenty years in their applica-
tion is timely, bearing in mind the long lead-time necessary
before damaging reaction becomes evident.
Issues for review
the classification of cement alkalinity and need for/
level of step changes between alkali classes
the alkali classification of composite cements and
cement combinations
the alkali contribution from aggregates
the reactivity and alkali contribution from recycled
aggregate other than recycled concrete aggregate
the classification of reactivity and alkali limit appli-
cable to aggregates
the limit of 0.60kg/m3 Na2Oeq alkali from other
sources
the appropriate publication medium for UK rules on
ASR.
Detailed review
The need for/level of step changes between alkalinity of
cements
Digest 330; Part 2 reports that investigations of UK struc-
tures affected by ASR have shown that damage is nearly
always associated with using an abnormally high alkali
cement (Na2Oeq >1.0%). The Concrete Society sub-com-
mittee report(5) found that concretes containing such high
alkali cements with normal reactivity aggregates showed
damaging ASR reaction at alkali levels of 5.2kg/m3 and
above. Laboratory tests on these cements and normal
reactivity aggregates have not reported deleterious, expan-
sive reaction below 4.8kg/m3.
Digest 330 goes on to record the successful use of
low alkali cement (Na2Oeq0.60%) to minimise damage
from ASR. It therefore classifies cements between the two
extremes as moderate alkali cements.
Digest 330 quotes its underlying principle as that more
precautions are needed to minimise the risk of damag-
ing ASR as the level of cement alkali increases. It gives
no rationale as to why no safety margin is applied to the
classification of cement as low alkali while, for its clas-
sification of cement as high alkali, applying a 25% safety
margin. There is no logic to this latter as this 25% margin
is applied in addition to limiting the overall alkali content
of concrete with high alkali cement with normal reactiv-
ity aggregates to 3.0kg Na2Oeq/m3, a further margin on
margin of approximately 50%. That is for a moderate
alkali cement, taking into account extreme variations from
the mean, the maximum expected alkali content would be
3.8kg/m3; the margin of 1.0kg/m3 against an assumed criti-
cal level of 4.8kg/m3, which equates to 26% of the maxi-
mum expected alkali content; the corresponding values for
a high alkali cement being 3.2kg/m3; 1.6kg/m3 and 50%.
The classification of low alkali cement is based
on alkali not exceeding 0.60% Na2Oeq/m3 for any spot
sample, while that for moderate and high alkali cements is
based on the declared mean alkali level. Basing a limit on
an indeterminate frequency of testing spot samples is ques-
tionable and open to dispute both statistically and from
results of testing samples taken on-site. It is suggested
that the basis of declared mean also be used for low alkali
cements and that the limit be adjusted to 0.50% Na2Oeq
CONCRETE July 09 33-48.indd 43
CODES AND STANDARDS
to allow for the variation around (above) the declared
mean. This would ensure equivalence, for all practical pur-
poses, to the former requirement for CEM I cement and
improved clarity of specification for low alkali composite
cements and combinations.
When Digest 330 was drafted and the limit for clas-
sification of high alkali cement was selected at >0.75%
Na2Oeq there were few UK CEM I Portland cements in
this category. Since then UK cement production has main-
tained this position by discarding alkali-rich dust. This
procedure is now less environmentally acceptable nor
sustainable on account of increasing restrictions on plac-
ing waste to landfill and in obtaining planning consent for
waste sites. It should also be noted that there are now no
certified low alkali CEM I Portland cements available on
the UK market.
There is some logic in maintaining the classification
of cement alkalinity since, should they become avail-
able, low alkali cements demonstrably impose little risk
whereas high alkali cements require more careful appli-
cation. Any imposition of a strict division between classes
automatically introduces critical comparisons at the
There is
boundary. The more technically correct approach might some logic in
be to use a sliding scale. However, this would be difficult maintaining the
to operate in practice and therefore it is suggested that the
step changes between classes be retained, while ensuring
classification of
that the risk of ASR is not increased. cement alkalinity
In this context, the criterion for classification of since, should
cement as high alkali needs to be reassessed. In the Irish
Republic, there is no such differentiation between moder-
they become
ate and high alkali cements yet they have successfully available, low
produced damage free concrete with a higher proportion of alkali cements
highly reactive aggregates than is generally encountered
in the UK. We have also seen an approach from Lithuania
demonstrably
to CEN/TC51(6) for the classification of low alkali impose little
cements as those having <1.0% Na2Oeq. Nevertheless, risk whereas
the UK approach allows clear specification for situations
when assessing the two risk elements: cement alkali and
high alkali
aggregate reactivity. There is no evidence that the alkali cements require
from high alkali cement is more of a risk than that from more careful
moderate alkali cement, even in combination with a
highly reactive aggregate; the Irish experience suggests
application.
that the overall alkali content of the concrete is the deter-
mining factor. Nevertheless, highlighting the alkali content
as high alerts the concrete producer to the need for care
in use.
Recommendation 1: That cements continue to be
classified as low, moderate and high alkali.
Recommendation 2: That the classification for reac-
tive alkalinity of all cements be based on the declared
mean alkali content and be: low alkali cements
0.50% Na2Oeq; moderate alkali cements >0.50
and <1.00% Na2Oeq; and high alkali cements
1.00% Na2Oeq.
The alkali classification of composite cements and
cement combinations
The requirements in Digest 330: Part 2: Table 1 classify the
alkalinity of composite cements (including combinations)
according to the alkali content of the CEM I type compo-
nent. It is noted that BS EN 197-1(7) NB.4.3 b) note 2, also
uses content. The principle used as the basis of this classifi-
cation should be reviewed to clarify that the criterion is the
contribution of alkali from each constituent rather than the
alkali content of one constituent.
Taken to the extreme, there is no technical basis for
considering a CEM III/B cement, consisting of 28% of a
CEM I type component with an alkali content of 1.01%
Na2Oeq, 2% gypsum and 70% blast-furnace slag, to be a
high alkali cement. The same would apply to a CEM IV
CONCRETE JULY 2009 43
23/06/2009 12:43:45
 CODES AND STANDARDS
fly ash cement. Consequently, the classification of alkalin-
ity of composite cements and combinations in Digest 330
requires revision.
Recommendation 3: That the alkalinity classifica-
tion of composite cements and combinations be
determined according to their declared mean alkali
content, calculated in accordance with BS EN 197-1:
national annex NB, Table NB. 2, and the alkali limits
set out in Recommendation 2 above.
The alkali contribution from aggregates
Digest 330 requires that alkalis from other sources (other
than cement) are taken into account against the overall
alkali content of the concrete. In discussion of the limit of
alkalis from other sources the alkali introduced as salt
from marine aggregates was considered. There has been
Increased no reported instance of other alkalis being released from
pressure to virgin UK natural aggregates, although it is well known
use a variety that many constituent minerals contain significant amount
of alkalis locked in their mineral crystal structure.
of recycled Recovered aggregates (the term used in BS EN 206-1(8)
and waste and previously termed reclaimed aggregate in BRE 330);
products, for that is natural aggregates recovered from the concrete
production process before the cement has set, are open
example glass to the risk that alkali from the mixing water or from any
or metallurgical admixture might have been absorbed into the aggregate.
Aggregates are normally used wet so that any alkali gained
slags, to
would be by slow diffusion rather than more rapid absorp-
minimise tion mechanism; also their use is limited to 5% of the total
demand on aggregate; therefore it is considered that the risk of signifi-
natural resources cant reactive alkali content is low.
Recommendation 4: That it should be explicitly
will result in stated in Digest 330 that for UK natural aggregates no
unfamiliar reactive alkali shall be considered to be contributed
other than that introduced by contamination, eg by
aggregates.
marine salt.
In the case of lightweight aggregate there can be sig-
nificant amounts of alkali bound within their structure.
There is no evidence that current UK lightweight aggre-
gates, either sintered fly ash or expanded clay, contribute
alkalis to cause damaging ASR in concrete. This may be
because the alkali remains locked within the structure or
that the porous nature of the aggregate provides sufficient
stress relief volume for any reactive gel formed. The basis
for the latter is indicated in the consideration of the effect
of air entrainment in the BRE Information Paper(3).
Recommendation 5: That it should be explicitly
stated in Digest 330 that no reactive alkali shall be
considered to be contributed from sintered fly ash or
expanded clay lightweight aggregates.
Other aggregates might release alkalis that contribute
to ASR reaction. Increased pressure to use a variety of
recycled and waste products, for example glass or metal-
lurgical slags, to minimise demand on natural resources
will result in unfamiliar aggregates. Such products might
contribute alkalis, other than marine salts, to the concrete.
The Concrete Bookshop
Visit: www.concretebookshop.com
44 JULY 2009 CONCRETE
CONCRETE July 09 33-48.indd 44
In these cases reliance on measuring chloride to deter-
mine the alkali contribution might not be adequate. A
RILEM method(9) has been developed and further inves-
tigation is necessary to confirm whether this can be cor-
related to assessment of reactive alkali in expansion levels
with calibrated mixes of known UK reactive aggregate
combinations.
Recommendation 6: That experimental work should
be undertaken to evaluate the RILEM method as a
reliable indicator of reactive alkali contribution from
aggregates.
Further information:
Part II of this article will appear in the August 2009 issue
of CONCRETE and will consider the guidance on recycled
aggregate; the classification system for aggregate
reactivity; the treatment of alkali arising other than from
cement and aggregate; and the possible routes for
publication of updated rules in the light of activities in
European standards committees.
Acknowledgements:
The author wishes to acknowledge the support provided
by the Mineral Products Association in undertaking this
review. Thanks are also due to those associated with
concrete production, concrete constituents production,
consultants, specifiers and the members of the Irish
Concrete Society for their input into this review.
* The author was chief chemist of Castle Cement, a
member of The Concrete Society Hawkins Committee and
chairman of its alkali limit working party. He is chairman of
BSI committee B/516 Cement and building limes.
References:
1. BUILDING RESEARCH ESTABLISHMENT, Digest 330:
2004. Alkali-silica reaction in concrete. Detailed guidance
for new construction. BRE Bookshop, 2004.
2. BUILDING RESEARCH ESTABLISHMENT, Digest 330.
Alkali-silica reaction in concrete. CRC Ltd, Watford, 1999.
3. BUILDING RESEARCH ESTABLISHMENT, Information
Paper 1/02: 2002. Minimising the risk of alkali-silica
reaction: alternative methods. CRC Ltd, Watford, 2002.
4. CONCRETE SOCIETY. Alkali-silica reaction: minimising
the risk of damage to concrete. Technical Report 30, The
Concrete Society, Camberley.
5. Concrete Society Sub-committee. Investigation of
structures affected by ASR. CONCRETE, Vol.31, March
1997, pp.25 27.
6. Lithuanian request under mandate 114, CEN Technical
Committee 51 Cement and building limes, Document
N838, 2007.
7. BRITISH STANDARDS INSTITUTION, BS EN 197.
Cement, Part 1: Composition, specifications and
conformity criteria for common cements. BSI, London,
2007.
8. BRITISH STANDARDS INSTITUTION, BS EN 206.
Concrete, Part 1: Specification, performance, production
and conformity. BSI, London, 2000.
9. RILEM/CANMET, FOURNIER, B. and LU, D. International
Centre for Sustainable Development of Cement and
Concrete (ICON/CANMET), Canada & College of
Materials Science and Engineering, Nanjing University of
Technology, China, Test Method for Alkali Release, 2003.
23/06/2009 12:43:46